In the T4 bacteriophage the primosome is formed by the helicase (gp41) and primase (gp61) complex. Here we use magnetic tweezers to manipulate a single tethered DNA hairpin. The substrate extension is a real-time reporter of the primosome-activity. We have investigated the action of the helicase in complex with primase or of it fusion protein on special DNA substrate having a small number of priming sites only in the lagging strand.Different primosome behavior is observed depending on (1) rNTPs mix and concentration, and (2) whether the helicase and primase are fused or not. In absence of rNTPs the experimental signal show increases in the substrate extension associated with dsDNA unwinding, and decreases in the substrate extension corresponding to the ssDNA rehybridization following helicase translocation on ssDNA. In presence of CTP no change is observed in the unwinding activity of gp41; however, the rehybridization signal is strongly altered by RNA primers synthesized by the primase. Our results suggest that, in these conditions, primase dissociates from the helicase during priming. On the contrary, the activity of the gp41/gp61 fusion protein shows different unwinding signal revealing the formation of ssDNA loops. Results are consistent with a “looping mechanism”, in which during priming helicase follows translocation and a ssDNA loop is formed and released after priming is over. Interestingly, in presence of all rNTPs looping can also be observed in the wild type primosome, even that much more unfrequently than in the gp41/gp61 fusion protein. Overall the results suggest that the primase processivity in the context of the primosome is very low. However, within the full replisome, the ssDNA looping might be the predominant mechanism which allows the helicase to continuously unwind the DNA without dissociate from the primase.