Suckling of a non-filial calf, or allosuckling, is an extreme case of allomaternal care in mammals. There have been many hypotheses postulated in an attempt to explain this behaviour; however, the supporting evidence differs, together with the conclusions drawn from the investigated variables. Previously, suckling bout rejection was analysed, and the milk theft and reciprocity hypotheses were both determined as the most appropriate explanation of allosuckling in giraffe. In this study, seven hypotheses were tested using different behavioural parameters, namely suckling bout frequency, suckling bout duration, and time spent suckling. It is well-documented that these parameters are associated with various aspects in ungulate biology; for example, suckling rejection typically reflects milk intake and parent–offspring conflict, whereas the suckling bout duration and frequency is associated with social behaviours (affiliation, bonding, social stress). In total, 22 nursing females and 47 suckling calves were observed, in four Czech zoos during a five-year period. The correlation of the observed parameters between the reciprocal female-calf dyads was found to be in line with the reciprocity hypothesis. In addition, non-filial calves tried to steal the milk from non-maternal females, supporting the milk theft hypothesis. Thus, the results support both the reciprocity and milk-theft hypotheses as the most plausible explanation of allosuckling in giraffe, and illustrates the importance of using suckling bout duration and frequency, and the time spent suckling, as behavioural parameters that may aid in explaining the extremity of maternal investment, such as allosuckling.