Inductive learning is of fundamental importance in our everyday lives and is enabled by our capacity to draw statistical inferences from limited sample observations to populations and vice versa. Contrasting former predictions, recent research revealed that great apes can flexibly draw inferences from populations to samples, suggesting that intuitive statistical reasoning is not restricted to humans. Here we investigated if apes truly engage in intuitive statistics by probing their abilities contrariwise: from samples to populations. In particular, apes were confronted with two covered cups containing populations of food items that differed in their relative frequency distribution of preferred and neutral food. Based on the observation of representative samples drawn from these populations, apes were allowed to choose between the two covered cups. In a series of five experiments we could show that apes reliably concluded from the samples to receive the preferred population, at least as long as they could rely on absolute frequencies of preferred food items. These results confirm the previous finding that great apes are capable of basic forms of intuitive statistics.