Objective and methods: In the Bonn Longitudinal Study (BLS), the course of neurologic development was analysed with regard to transient abnormal neurologic signs (TANS). Abnormal neurologic signs (ANS) are defined when present in at least one developmental area at one examination. From birth to adulthood (mean 23 years) 108 preterm infants, 81% of very low birth weight (VLBW) were examined with regard to their psychological, neurologic and physical development: 62 appropriate for gestational age (AGA), 46 small for gestational age (SGA), 27 with postnatal catch-up growth of head circumference (group A) and 19 without catch-up (group B); 73 full terms served as controls. The dropout rate was 7.6%. Results: ANS showed a great inter- and intraindividual variability. Episodes of neurologic abnormalities changed with those of normality. ANS reappeared when new abilities developed. ANS were observed mainly during the first year. The incidence was higher in preterms (AGA 49%, SGA A 63%, SGA B 61%) than in full terms (15%). Transient ANS (TANS) were diagnosed in 42% of AGA, 63% of SGA A, 33% of SGA B preterms and 12% of full terms. The recovery rate, i.e., the TANS to ANS percentage, was higher in AGA (86%), SGA A (100%) and full terms (82%) than in SGA B (55%) preterms. Conclusion: It is impossible to predict early on whether ANS will be transient, i.e. TANS. Longitudinal analyses are needed. The outcome depends on recovery from or persistence in ANS. (C) 2000 Elsevier Science Ireland Ltd. All rights reserved.