• A quarter of the children whose parents had separated said no one talked to them about the separation when it happened. Only 5 per cent said they were given full explanations and the chance to ask questions. Most reported that they were confused and distressed by the separation.
• Grandparents and friends were children's key confidants in the weeks following separation; confiding in fathers and siblings was rare.
• Those children who felt they had poor relationships with their parents and that they were more involved in conflict between parents and step-parents tended to have more adjustment problems.
• Children whose parents had experienced early adversities, teenage pregnancy and several changes in relationships described their relationships with their parents as less warm, confiding and companionable than other children.
• Children who felt close to their maternal grandparents had fewer adjustment problems. No such association was found for closeness to paternal grandparents.
• Over half the children who lived in two households because of separated parents were positive about their 'divided' lives. Those who had an active role in decisions about these arrangements and those who said they were able to talk to parents about their problems concerning their 'divided' lives were more likely to have positive feelings about moving between households.
• The views of children as young as five (obtained using drawings and family 'maps') were similar to the verbal accounts given by older children.

Increasing numbers of children experience the separation of their parents; this separation frequently forms part of a chain of changes in their families, with new partners for their parents, periods living with a single parent, sharing their homes with stepsiblings and half-siblings, and the experience of living in two households. These events increase the risk that children will have problems in adjustment, with their health and at school. But the way children respond to family change varies greatly. This research investigated the views of children on their experiences and what they found helpful.

In the weeks following parental separation, the key people in whom children confided intimately were grandparents - especially on their mother's side - and friends; the next most frequent confidantes were mothers. Children rarely confided in their fathers or siblings.

Children differed very much in the extent of conflict and hostility they described in their relations with their parents, but these differences were not linked to whether the family was a stepfamily, single parent or 'birth' family. They were associated with adjustment problems: children who had difficult, negative relationships with either father or mother were more likely to be anxious, depressed, worried or to be aggressive or have problems at school. It was not clear whether the poor relationship caused the adjustment problems or vice versa.

The children completed 'maps' of the emotional closeness of their various relationships and the youngest (aged 5-6) also drew pictures of their families. Both 'maps' and drawings paralleled the findings from talking to the children. For instance, children who placed their fathers or stepfathers on the 'map' as not emotionally close were almost three times more likely to be reported by teachers as having problems at school.

The children's views of their relations with their parents - whether from interviews or 'maps' - were more negative in families in which the parents themselves had experienced adversities in their earlier lives, had been teenage parents, or had had several changes in relationship. Strikingly, the children's view of their closeness to their fathers (birth or step-fathers) predicted which children would have adjustment problems even when these aspects of the parents' lives and other features of the family situation, such as income, were taken into account.

Many children said they missed their non-resident parent very much indeed and longed to see more of them. They made practical suggestions on how to achieve this. One child commented that he would like to see his father at weekends rather than on weekdays, because they did not get time to talk on schooldays. Others said they wanted to do things with their non-resident parents, not simply watch TV. The unreliability of non-resident fathers was a common theme and caused much distress.

Over half the children who lived in two households because of separated parents were positive about their 'divided' lives. Some, for example, were pleased to get away from their step- and half-siblings at the weekend. Children who had been given a role in decisions about visiting times etc were more likely to feel positive about moving between two households.

If children felt that they could not talk about their problems in moving between two households or problems they might be experiencing in their 'other' households, they were more negative about their 'divided' lives. However, they were very sensitive to criticism of one parent by the other and to conflict between birth parents.

The children saw the role of step-parents in very different ways. Some stressed that the step-parent should be a friend, others said a parent. Many said they found discipline by a step-parent difficult to take. Fifty per cent felt that they took second place to children born to their own parent and step-parent and 30 per cent felt they were displaced by their step-parents' own children.

Children who felt emotionally close to their maternal grandparents were less likely to have adjustment problems - to be depressed, anxious, worried or aggressive and difficult at school. This effect was significant even when other family difficulties such as mothers' depression or problems in the child's relationships with the parents were taken into account.

• Affectionate supportive friendships were more frequent among children whose relationships with their mothers were warm and close.
• Children frequently involved in conflict between their parents (including step-parents) were less likely to have affectionate supportive friendships.
• However, children living with, but not getting along with, stepmothers had more extensive contact with friends than did other children.
• The 'maps' drawn by the 5- to 7-year-olds confirmed the picture from the interviews with older children: emotional closeness to mothers was associated with emotional closeness to friends.

• The extent to which children felt they didn't know what was happening within the family and the fact that they may interpret the situation as meaning they are no longer loved by the parent who has left.
• The low level of children's communication with fathers.
• Children who reported frequent shared family activities in their main family household had fewer adjustment problems.
• Grandparents and friends were key sources of support for children, and relations with grandparents were linked to children's adjustment.

The children talked about their current family relationships with parents, step-parents and siblings, about their experiences at the time their parents separated and about their friendships. They also completed 'maps' showing the people to whom they felt emotionally close and the younger children (aged 5-7) drew pictures of their families. Parents and teachers completed reports on the children's adjustment.