Kenya witnessed an upsurge in early childhood mortality since the late 1980s through the 2003 period. The levels of infant mortality rates increased from 59 deaths per 1000 live births in 1989 to 78 deaths per 1000 live births respectively by 2003 (CBS, et al., 2004). These trends represent a percentage increase of about 32 percent in infant mortality and 29 percent in child mortality for the period of 1989-2003. Scanty research exists on the probable factors behind the upturn in infant and child mortality during the 1988-2003 period. One clearly investigated issue is that the upturn in infant mortality is not as a result of measurement error or deteriorating data quality (Hill et al., 2001; Wangila et al., forthcoming). This paper applies the Weibull regression models in accelerated failure time (AFT) framework using Stata 10 (StatCorp, 2009) to investigate some of the probable factors that may have fuelled the upsurge in infant mortality. The Mosley and Chen (1984) conceptual framework is used as a guide in the identification of the covariates.