Some remarks on a conjecture of de Oliveira