SummarySomatic embryogenesis was induced in immature zygotic embryos of guava (Psidium guajava L.) at 10-weeks post anthesis using 2 – 38 d pulses and continuous treatments with different concentrations (0.001 – 3.0 mg l–1) of 2,4-dichlorophenoxyacetic acid (2,4-D). Subsequent development of somatic embryo was observed by sub-culturing onto plant growth regulator (PGR)-free, full-strength Murashige and Skoog (MS) agar medium containing 3% (w/v) sucrose. The maximum responses were shifted gradually from 0.01 mg l–1 2,4-D (continuous or 38 d treatment), to 0.05 mg l–1 2,4-D (28 – 18 d treatment), to 0.1 mg l–1 2,4-D (14 – 12 d treatment) and, finally, to 0.5 mg l–1 2,4-D (10 – 8 d treatment).The highest frequency of embryogenesis (68.8%) and embryogenic intensity (69.2 embryos per explant per culture), as well as moderate frequencies of convertible elongated and short torpedo-stage somatic embryos (15% and 42.9%, respectively), with the highest efficiency (27.5) achieved in the 8 d treatment compared to continuous treatment of zygotic embryo explants with 0.5 mg l–1 2,4-D in MS agar medium containing 3% (w/v) sucrose at both the induction and development phases. Cellular pools of total, free, conjugated, and bound forms of various polyamines (e.g., putrescine, spermidine, and spermine) showed slow increases in the presence of 2,4-D during the 8 d induction phase. This was followed by rapid increases immediately after sub-culturing explant tissue onto the PGR-free MS development medium. The concentrations of various polyamines declined gradually over time on MS development medium. The production, metabolism, and use of different forms of various polyamines at various times (d) after culture initiation correlated with their temporal regulation during the induction and development phases of somatic embryogenesis. Exogenous application of different polyamines during the induction phase resulted in slightly increased frequencies, intensities, and efficiencies of embryogenesis, but with no observable change in the frequencies of the various categories of somatic embryos. The present study indicated that the process of somatic embryogenesis in guava may be less sensitive to PGRs and was induced even in the continuous presence of 2,4-D. Furthermore, somatic embryogenesis in guava was temporally regulated as a function of the exogenous concentration of 2,4-D. In addition, the temporal regulation of somatic embryogenesis was also associated with high levels of production, metabolism, and use of cellular polyamines during the different phases. Hence, improvements in the process of somatic embryogenesis in guava could be regulated by timed applications of 2,4-D alone, or in combination with polyamines, thereby modulating endogenous levels of the latter.