This study assessed the reliability of electricity supply to industrial customers in Uganda, using selected industrial areas as a case study. A quantitative design was adopted, and a coded dataset of 100 industrial connections was analysed using descriptive statistics, Pearson correlations, multiple linear regression of monthly outage hours and binary logistic regression of significant output loss, supported by diagnostic tests. The dataset was simulated for study purposes and was designed as a template to be replaced by real utility and firm records. The study results indicated that customers experienced an average of 8.8 outage hours per month, and that a dedicated feeder (B = -5.69 hours, p < 0.001) and more preventive maintenance reduced outages, while older feeders and greater distance from the substation increased them. A dedicated feeder reduced the odds of significant output loss (odds ratio = 0.22, p = 0.008), while greater distance from the substation increased them (odds ratio = 1.24, p = 0.014). Heteroscedasticity was detected and addressed through robust standard errors, which left the conclusions unchanged. The study concluded that reliability depended chiefly on network configuration, age and maintenance, and it recommended dedicated feeders for industrial parks, refurbishment of ageing lines, structured preventive maintenance and publication of reliability data.