This is a retrospective analysis of a multicentre randomised controlled trial (RCT) where we concluded that CeasIng Cpap At standerD criteriA (CICADA) in premature babies (PBs) <30 weeks gestational age (GA) was the significantly better method of ceasing CPAP. To identify factors that may influence the number of attempts to cease CPAP, we reviewed the records of 50 PBs from the RCT who used the CICADA method. PBs were grouped according to number of attempts to cease CPAP (fast group ≤2 attempts and slow group >2 attempts to cease CPAP). There were 26 (fast group) and 24 (slow group) PBs included in the analysis. Results showed significant differences in mean GA (27.8 ± 0.3 vs 26.9 ± 0.3 [weeks ± SE], p = 0.03) and birth weight ([Bwt]; 1080 ± 48.8 vs 899 ± 45.8 [grams ± SE], p = 0.01) between groups. Significantly fewer PBs in the fast group had a patent ductus arteriosus (PDA) compared to the slow group (5/26 (19.2%) vs 13/24 (54.2 %), p = 0.02). Bwt was a significant negative predictor of CPAP duration (r = -0.497, p = 0.03) and CPAP ceasing attempts (r = -0.290, p = 0.04).PBs with a higher GA and Bwt without a PDA ceased CPAP earlier using the CICADA method. Bwt was better than GA for predicting CPAP duration and attempts to cease CPAP.Our previous studies showed that CeasIng Cpap At standarD criteriA (CICADA) significantly reduces CPAP time, oxygen requirements and caffeine use. Some PBs however using the CICADA method required >2 attempts to cease CPAP ('slow CICADA' group).PBs in the 'fast CICADA' group (<3 attempts to cease CPAP) (a) have longer gestational age and higher birth weight, (b) shorter mechanical ventilation and (c) lower incidence of patent ductus arteriosus. Attempts to cease CPAP decreased by 0.5 times per 1 week increase in GA and 0.3 times per 100-g increase in birth weight for PBs <30 weeks gestation.