The opinion of the Court was delivered by HANDLER, J.

This appeal raises the same issues and invokes the same concerns as the companion case, In re J.C., 129 N.J. 1 , 608 A.2d 1312 (1992). In this case, the mother voluntarily placed her newborn infant with the Division of Youth and Family Services (DYFS or Division or agency) for temporary custody and care. Following a long period of separation of mother and child the Division brought an action to terminate the parental rights of the mother in order to prepare for the permanent placement of the child. The trial court refused to terminate the mother's parental rights, concluding that the mother had not abandoned the child and was fit to raise the child, and, further, that the child would not suffer psychological harm in being removed from her foster parents and returned to her mother. The Appellate Division affirmed. The Court granted the petition for certification. 127 N.J. 549, 606 A.2d 362 (1991).

I

B.F. became pregnant with K.L.F. after having been gang-raped in New York City. She came to New Jersey to have her child, believing that New York City was too dangerous a place in which to raise children. She gave birth to a healthy daughter at Hackensack Hospital in Bergen County in November 1988.

At the time she gave birth, B.F. was homeless. Unable to find shelter for herself and her daughter she entered into a temporary custody agreement with the Division of Youth and Family Services and consented to the temporary placement of K.L.F. in foster care. N.J.S.A. 30:4C-11. The agency arranged for B.F. to live at the Bergen Shelter, a facility that did not allow infants or children.

B.F. visited with K.L.F. twice during December at the DYFS offices in Hackensack. When B.F. came for a third visit, she was unable to see K.L.F. because the social worker handling the case was sick. Shortly thereafter, B.F. left the shelter, and returned to New York City in search of permanent housing.

During the year and a half that followed, B.F. had neither work nor a home, and lived in shelters and with friends. At trial, she testified that she called DYFS from pay phones in New York numerous times during that period but was unable to reach an agency worker who knew about K.L.F.'s case.

In the meantime K.L.F. lived in the home of foster parents. DYFS caseworkers sent letters to various State agencies and made phone calls but were unable to communicate with B.F. In May 1990, eighteen months after K.L.F. had first been placed, DYFS concluded that she needed a permanent home and moved her to a new set of pre-adoptive foster parents with whom she currently lives.

Less than a month later B.F. reached K.L.F.'s new case worker at DYFS, and requested to see her child so that she might eventually regain custody. DYFS informed B.F. that the agency was bringing a legal action for guardianship and that she would have to wait for a court to decide whether visitation should occur. DYFS also told B.F. that she would have to release medical information and later undergo psychiatric evaluation before visitation could be permitted. Because the agency refused to allow her to visit her child, B.F. concluded that she had to go to court to obtain visitation. She traveled from Staten Island to the Bergen County Courthouse in Hackensack to file a pro se motion for visitation. Although Bergen County was the correct venue, employees at the courthouse sent her to the Passaic County Courthouse in Paterson. When B.F. finally arrived in Paterson, she was not permitted to file her motion there and was directed to return to Bergen County. Frustrated, B.F. returned to Staten Island without having filed a motion for visitation or otherwise obtained assistance.

B.F. thereafter did find a stable living situation, signing a three-year lease on an apartment in Staten Island where she now lives with another daughter. Nonetheless, on March 23, 1991, DYFS petitioned for guardianship and termination of parental rights based on abandonment and the best interests of the child. N.J.S.A. 30:4C-15. At that time K.L.F. had been living with her current foster parents for ten months. DYFS concluded that K.L.F. had bonded with them and that moving her again would cause her psychological and emotional harm.

At the trial both parties introduced evidence chronicling the foregoing events primarily on the issue of whether B.F. had abandoned her child. In addition, both parties presented expert testimony relating to B.F.'s parental fitness and to harm to the child resulting from separation from her foster parents.

II

The legal framework within which the issues must be resolved is set forth in In re J.C., supra, 129 N.J. 1 , 608 A.2d 1312. As in that case, the mother, B.F., voluntarily agreed to place her child in the temporary custody of DYFS. Under the voluntary-placement scheme, the initial decision to place a child in foster care rests solely with the child's parent or guardian. However, before actually providing foster care services DYFS must itself determine that a child's welfare is endangered and that the child's needs cannot be met either through financial assistance or placement with family or friends. N.J.S.A. 30:4C-11. To obtain guardianship over children who have been placed in foster care, the agency must bring an action to terminate the natural parents' rights under N.J.S.A. 30:4C-15 (section 15). In this case, DYFS proceeded under subsections c and d of section 15, which authorize it to seek guardianship respectively when that is in the best interests of the child and when the parent has abandoned the child. Guardianship cannot be awarded in an action brought pursuant to section 15 unless the court itself determines that it is in the child's best interest under N.J.S.A. 34:4C-20.

Those broad statutory standards were explained by this Court in New Jersey Division of Youth and Family Services v. A.W., 103 N.J. 591, 512 A.2d 438 (1986). We there noted that the statutory scheme must balance the constitutional rights of parents against the interests and rights of their natural-born children. Id. at 599-60, 512 A.2d 438. The statute requires DYFS to make an affirmative demonstration that the child's best interests will be "substantially prejudiced" if parental rights are not terminated. Id. at 603, 512 A.2d 438. That demonstration ordinarily requires clear and convincing evidence of serious impairment of the child's health or development caused by the parent, the inability of the parent to rectify or overcome that kind of harm, the detrimental effects of delay in arranging for permanent placement of the child, and the absence of alternatives to termination of the parental rights. J.C., supra, 129 N.J. at 8-9, 608 A.2d 1312 (summarizing A.W.).

We also pointed out in J.C. that the Legislature, in 1991, amended N.J.S.A. 30:4C-15, the statute governing agency action in the termination of parental rights. L. 1991, c. 275, § 7. Those amendments adopted substantially the standards in A.W. Reflecting an important societal goal of preserving the natural family as a unit, the statutory scheme requires a finding that the natural parents would themselves do substantial harm to the child if parental rights are not terminated and the child is returned. The burden falls on the State to demonstrate by clear and convincing evidence that the natural parent has not cured the initial cause of harm and will continue to cause harm to the child. J.C., supra, 129 N.J. at 10-11, 608 A.2d 1312.

III

Following the trial, the court determined that B.F. had never abandoned the child. It found that B.F. had consented to the placement of K.L.F. because she believed it was in the child's best interest, but that she had done so under the impression that she would be reunited with the child in a shelter facility that allowed mothers to stay with their children. According to the court, B.F. had not been aware of the "tortuous legal procedures through which she would have to go" to regain custody of the child.

The court found that after she had placed K.L.F. with DYFS, B.F. had no income and no home and was therefore unable to receive telephone calls or the many letters D.Y.F.S. sent her. The court found that she had made "numerous" and "persistent," albeit "ineffectual, paltry, and meager" attempts to communicate with DYFS in order to see the child and that those efforts included phone calls to DYFS on numerous occasions. Specifically, the court found that no phone calls were separated by more than a year and that "no effort to communicate with the Division or to see the child were separated by more than a year."

In further support of its finding that B.F. had not abandoned the child, the court noted that after she had placed K.L.F. in foster care, she began to prepare for life with her children. The court credited B.F. with rehabilitating her "lifestyle" and with finding and furnishing an apartment in a neighborhood conducive to raising children. It also emphasized that she had worked persistently to regain custody of her child, including a frustrated effort to file a pro se motion for visitation first at the Bergen County Courthouse and then at the Passaic County Courthouse. Based on those findings, the court concluded that the "evidence of abandonment [was] meager and [was] substantially outweighed by the commendable efforts on the part of [B.F]."

The Appellate Division determined that the record "amply support[ed]" a finding that B.F. had not abandoned her daughter. It reasoned that "[p]erhaps a person of greater fortitude and sophistication could have overcome those obstacles; but the fact that B.F.'s efforts were, as found by the trial judge, only `meager' and `ineffectual' does not indicate an abandonment of the child" (quoting A.W., supra, 103 N.J. at 615, 512 A.2d 438).

We are satisfied that the lower courts correctly concluded that B.F. did not abandon her child. Abandonment requires a finding that parents, although physically and financially able to care for their children, willfully forsook their parental responsibilities. See N.J.S.A. 30:4C-15(d). The concept of abandonment entails a willful surrender or intentional abdication of parental rights and duties. See In re Adoption of Children by D., 61 N.J. 89 , 94-95, 293 A.2d 171 (1972) (finding forsaken parental obligation "require[s] a past course of conduct amounting to intended abandonment or very substantial neglect of both parental duties and claims, with no reasonable expectation of any reversal of that conduct in the near future"). A lack of communication or planning that is not the fault of the parent does not equate with parental unfitness, nor does it constitute abandonment for statutory purposes.

The statute itself attempts to forestall the abandonment of children. It encourages the maintenance of relationships between biological parents and their child even though under the care of foster parents. See, e.g., N.J.S.A. 30:4C-15(d) (before agency can find abandonment, it must make diligent efforts to encourage and strengthen the parental relationship); see also N.J.S.A. 9:6B-1 to -6 (Child Placement Bill of Rights gives children in foster care right to visit with natural parents and natural siblings on regular basis). In effect, the Division must take affirmative steps to avoid the eventuality of abandonment, particularly when, as in this case, the parent expressed a sincere interest in caring for her child.

Furthermore, under the voluntary agreement by which K.L.F. was placed, B.F. had the right to take back custody of her child at any time. If the agency wanted to retain custody despite her request, it was required to seek a court order authorizing it to do so. N.J.S.A. 30:4C-12. To the extent the Division's decision to refuse B.F. custody or contact with her child was based on its doubts about B.F.'s parental fitness, it was obliged by law to obtain a court order authorizing its own continued and exclusive custody over the child. DYFS was not empowered unilaterally to displace B.F. as a parent without judicial approval. See N.J.S.A. 30:4C-12 (conditioning order taking custody of child on finding that "parent . . . is grossly immoral or unfit."); N.J.S.A. 30:4C-58 (requiring periodic review by Child Placement Review Board to determine whether out-of-home placement is in the child's best interests); see also Santosky v. Kramer, 455 U.S. 745, 753, 102 S.Ct. 1388, 1395, 71 L.Ed.2d 599, 606 (1982) (constitutional protection afforded to rights of natural parents is not diminished by fact that "they have not been model parents or have lost temporary custody of their child to the State").

When viewed in their entirety, the facts indicate that DYFS erred in concluding that B.F. had abandoned her child. The lower courts correctly rejected that determination.

IV

The standard invoked by the lower courts to determine whether the parental rights of B.F. should be terminated was whether she would cause harm to her child. The courts below applied that test in order to evaluate the psychological harm to the child that might result from separating her from her foster parents. K.L.F. came to her pre-adoptive parents after having lived with another set of foster parents for eighteen months. Much of the testimony explored the significance of the change between her first and second foster parents.

Dr. Martha Page, testifying on behalf of the Division, acknowledged that K.L.F. had apparently adapted successfully to the change in caretakers at eighteen months. Nevertheless, Dr. Page expressed concern over the cumulative effect of another change in K.L.F.'s life. Any further disruption in the child's life "[would] expose her to emotional risks that I would prefer not to see her exposed to" and "could introduce stress that I would prefer not to see introduced into her life." Dr. Page suggested that the effects of breaking the bond "could be permanent" and "would be long term time bombs." Dr. Page found also that K.L.F. had "a very mutually warm and loving interaction" with her pre-adoptive foster parents and that she would suffer psychological harm if she were separated from them and reunited with B.F.

As a basis for her conclusions, Dr. Page expressed her general views on the impact of disruptions and changes on child development. Children, she stated, need consistency and continuity in care. Children who are transferred many times spend their energy and efforts acclimating, and as a result fall below other children in development. Moreover, "a child can learn that the world is a safe and predictable place only if he or she is with [a] safe and predictable environment." If psychological bonding fails to occur, children become distrustful, are often disruptive, have learning difficulties, and are unable to form lasting relationships.

Dr. Jerome Goodman testified for B.F. He concluded that the detrimental effect of removing K.L.F. from her foster parents and placing her with her mother would be minimal. He explained that "the vulnerability of a child in moving is very concentrated between nine months and eighteen months because this is the time in which the child generally learns who is stranger and who is home base." According to Dr. Goodman, if the child has successfully bonded with a primary caretaker during the critical early years of her life, in this case during the foster placements, the child acquires a set of adaptive skills that allow her to absorb significant changes successfully.

The trial court accepted the testimony of Dr. Goodman that K.L.F. could readily overcome any problems associated with leaving her pre-adoptive parents. The court emphasized that Dr. Goodman's views were supported by the Division's own records. The court noted that those records "clearly and unambiguously articulate that when this child was moved from the first custodial family to the present so-called pre-adoptive family, her reaction and her adjustment was marvelous, was appropriate, was normal, was suitable." It was persuaded by Dr. Goodman's analysis that K.L.F.'s successful experience of being transferred from one set of foster parents to another indicated that she would similarly adjust to being returned to her natural mother.

Both experts also testified about B.F.'s fitness as a parent as an aspect of the assessment of harm to the child on being reunited with her natural mother, and specifically whether B.F. had the capability to overcome such harm. Again, the experts disagreed.

Dr. Page reported that B.F. was "impulsive" and "grandiose in her estimates of what she could do." Dr. Page believed that she might not perform well under stress and that she still had a considerable way to go in her "recovery" from chronic homelessness and lack of planning. Dr. Page also reasoned that B.F.'s personality might lead her to engage in overindulgence or excessive gratification in the care of her children.

Dr. Goodman concluded that B.F. was a fit parent. Although she was depressed, the depression was situational, related to the circumstances of her life, especially the conflict with DYFS about custody over K.L.F. Rather than having unrealistic expectations or a grandiose personality, B.F. had a strong and sincere motivation to raise her child. Dr. Goodman stated that he could find no mental illness or history that rendered B.F. unable to raise K.L.F.

The court refused to accept the testimony of Dr. Page, finding it distorted by personal judgements and biases. Based in part on her demeanor and her unwillingness to concede anything positive about B.F., the court concluded that Dr. Page's examination of B.F. was neither impartial nor credible. The court also faulted Dr. Page for discounting the significance of the difference between the child's racial background and that of the pre-adoptive parents and for giving too little weight to the fact that K.L.F. had adjusted successfully when she was moved from the foster home to the pre-adoptive home.

The Appellate Division similarly found sufficient evidence in the record to support the finding that B.F. was a fit parent with the ability and capacity not only to care for K.L.F. but also to minimize the trauma involved in separating K.L.F. from her foster parents. The Appellate Division was satisfied that there was not clear and convincing evidence that returning K.L.F. to her natural mother would cause serious psychological harm. The court determined that even if Dr. Page's testimony were credited and accepted, it was not sufficient to support the agency's burden of proof.

In short, Dr. Page found K.L.F. had handled with substantial success two major moves in her young life, but she was concerned about the child's ability to respond well to yet another move. That is a legitimate concern: surely it would be preferable for a child not to be so disrupted. But the fact that another such move inevitably carries some risk is not clear and convincing evidence that the return of the child to her mother would be "so injurious" as to justify terminating the mother's rights. A.W., supra, 103 N.J. at 605, 512 A.2d 438.

We agree with the resolution of that issue by the lower courts. As we noted in J.C., supra, 129 N.J. at 18, 608 A.2d 1312, injury to children need not be physical to give rise to State termination of biological parent-child relationships. Serious and lasting emotional or psychological harm to children as the result of the action or inaction of their biological parents can constitute injury sufficient to authorize the termination of parental rights. Ibid.

We are acutely mindful of the controversy that surrounds the competing psychological theories of the effects of bonding. Experts differ over the assumptions relating to the adaptability of young children. As we observed in J.C., the court has a responsibility to make sense of the competing views presented by the experts and to assure a complete and balanced presentation of all relevant and material evidence sufficient to enable it to make a sound determination of the child's best interests. Id. at 20-21, 608 A.2d 1312.

Here, the trial court's approach was calculated to reach a balanced and informed resolution of the evidence in light of the contrasting views of the experts concerning the psychological harm that might befall K.L.F. Dr. Page's belief that a transfer in custody would cause harm was based on her assumption that change itself was harmful. However, the record supports the contrary conclusion of the trial court based on Dr. Goodman's beliefs that after a certain age children can absorb change, particularly if it takes place under the right circumstances. Dr. Page also saw problems in the combination of the child leaving a stable foster home and returning to a mother who might not have the ability to care for the child adequately. However, the record indicates that B.F. has undertaken substantial steps to prepare a stable and secure home environment for her daughter. In addition, the fact that B.F. is currently living with and caring for K.L.F.'s younger sister indicates that B.F. is not only a fit parent but will also be able to cushion the stress that will inevitably accompany the return of K.L.F. The evidence presented by DYFS did not support or sustain a conclusion that moving children under those circumstances will, to a reasonable psychological certainty, cause serious harm, especially given the contradictory testimony by Dr. Goodman. The standard is not that the end result cause no pain or trauma but that the child be kept from its parents only to avoid serious and lasting harm. See J.C., supra, 129 N.J. at 10-11, 608 A.2d 1312; In re Guardianship of J.E.D., 217 N.J. Super. 1 , 524 A.2d 1255 (App.Div. 1987); see also Division of Youth Family Servs. v. T.C., 251 N.J. Super. 419 , 440, 598 A.2d 899 (App.Div. 1991) ("We cannot . . . reject the thesis that in particular circumstances, foster-parent bonding may justify parental termination in order to spare the child grievous and irreparable psychological harm.").

We noted in J.C. that risks inhere in the use of bonding and psychological-parental theories. The theories are abused if routinely relied on to keep children in foster care rather than return them to their parents. 129 N.J. at 20, 608 A.2d 1312. If child-placement decisions are based solely on the need for continuity in care the results can be especially harsh on parents with few resources, who should not be put under siege by social welfare agencies. Ibid. The facile use of the bonding theory can increase the risk of institutional bias militating in the direction of permanent placement and adoption of children in foster care. Ibid. Here, for example, the trial court was troubled over what it viewed as hostility toward B.F. by the agency's expert. The court concluded that the expert's testimony crossed the line from scientific analysis of the harm facing K.L.F. to personal judgment about the character of B.F. In a process in which stability and continuity are made paramount, such subtle prejudices may unfairly weigh the process against parents with fewer material resources.

In a case such as this the risk comes perilously close to the reality. We are compelled to note that much of the bonding that has taken place in this case could have been avoided if the agency had correctly followed its mandate to use due diligence and its best efforts to reunite children with their natural parents. N.J.S.A. 30:4C-15; 30:4C-58. When B.F. requested that K.L.F. be returned to her custody, the child had been with her current foster parents for only a month. When DYFS petitioned for guardianship in March 1991, the child had been with the foster parents for ten months. Regrettably, litigation has extended that period even more. By encouraging her foster parents to believe that K.L.F. was on the way to becoming their child, and to view their interests and those of the child as being opposed to her reunification with her biological parent, DYFS may have increased the amount of bonding that has occurred. That those in the child welfare system not tip the scales and encourage a foster parent-child bond to develop when the natural parent is both fit and anxious to regain custody is essential. Indeed, we suspect that if the agency had allowed visitation and begun a process of reuniting B.F. with her daughter, it could have helped create a bond between the daughter and her mother that would have greatly mitigated any harm from being removed from foster parents. T.C., supra, 251 N.J. Super. at 437, 598 A.2d 899. However, we are satisfied that the record fairly supports the conclusion that the harm in returning K.L.F. to her mother is not extensive or irremediable.

V

We conclude that the trial court and Appellate Division applied the appropriate legal standards to the facts. They correctly determined that there was no clear and convincing evidence of either willful abandonment or serious and lasting harm to the child to justify the termination of parental rights. Because the record does not justify termination on either ground, the judgment of the Appellate Division is affirmed.

The trial court originally ordered DYFS to arrange for a series of overnight visits leading to a prompt transfer of custody. We remand to the trial court to implement that order.