<p>Cassava (<i>Manihot esculenta</i> Crantz) productivity among small-scale farmers remains low, primarily due to the use of low-yielding landraces and limitations in multiplication of planting materials of improved varieties. This study evaluated the field performance of cassava plantlets derived from leaf buds compared to conventional stem cuttings. A split plot experiment designed in a randomized complete block design (RCBD) with three replications was used to evaluate the field performance of five improved varieties of cassava, namely Kizimbani, Mkombozi, Mkumba, TARICASS 4, and Mkuranga 1, on growth and yield. The cassava propagation methods tested included leaf buds from upper, middle, and lower branch positions and conventional stem cuttings. Data collected included plants survival rate in 30 days after field establishment, average plant height (cm), stem thickness (mm), number of cuttings, shoot fresh weight (kg), number of roots per plant, root fresh yield (t ha^− 1), and dry matter content (DMC%) of roots. The results showed significant (<i>P</i> &lt; 0.001) differences for both variety and plantlet origin. Among the varieties tested, TARICASS 4 was a superior variety, with the highest shoot weight (6.5&#xa0;kg plant^− 1), root fresh yield (34.01t ha^− 1), and DMC (43.62%). Regarding the leaf bud plantlets, the plantlets derived from the upper position demonstrated superior performance compared to those from the middle and lower positions. Conventional cuttings consistently outperformed leaf bud-derived plantlets in terms of plant height, stem thickness, and yield. However, the upper-position leaf-bud-derived plants showed growth and yield potential similar to the conventional cuttings. This offers a sustainable, viable alternative method of cassava multiplication, particularly when mass multiplication is desired and where few materials are available, especially for research.</p>