This paper presents an adaptive maximum power point tracking (MPPT) algorithm. The aim is to dynamically adjust the step length for updating duty ratio (or operating voltage) so as to make full utilization of the output power of photovoltaic (PV) systems, even under the rapidly changing atmospheric conditions. To this end, the average slope in terms of voltage and power is exploited for reducing the harmful effect of noise and error (incurred in measurement or quantization) on the slope. Also, a statistical decision-making scheme is employed for reliably deciding the time instant at which atmospheric conditions actually change. Empirical study has adduced grounds for its dominance over existing references.