On the 18th of April 2010 the first phase of the eruption came to an end and the second phase started. The activity of the eruption changed from being a phreatomagmatic eruption to a strombolian eruption. Strombolian eruptions are relatively low power eruptions and characterized by both lava flows and ejection of tephra away from the eruption site by small explosions. The larger tephra fragment will start to build up a cinder cone (crater) but finer tephra is carried away from the eruption site in a low lying plume. The explosions happen due to the buildup of gas pressure in the uppermost part of the lava in the crater, and it can take anywhere from seconds to minutes between each explosion.

The reasons for the change in the activity are that the ice around the eruption sites had been melted away and water was no longer able to enter the vents, and that the erupting lava became less silicic, both of which cause reduced explosive activity. The power of the eruption decreased during this phase as well, but the flow of lava was still estimated at 20-50 tons per second. During this phase the eruption was classified as a mixed eruption, with neither tephra nor lava production constituting more than 95% of the erupted material. The ash produced during this phase was much coarser than from the previous phase and was also not carried as high up into the atmosphere with the plume. The eruption plume from the eruption became smaller and reached much lower into the atmosphere than during the first phase, and as tephra production had decreased the column became much lighter in color. 

The fissure that opened on the 15th of April became the main hub of activity, and remained until the end of the eruption, and a spatter cone was formed. The lava produced during this phase extended to about 3 km away from the spatter cone and flowed down and through the outlet glacier of Gígjökull. A large canyon was formed in the ice there due to ice melt from the lava, and the outlet glacier has yet to recover from this. The ice melt produced large quantities of water which entered the river of Markarfljót, but it did not cause large scale flooding.

The second phase lasted until the 4th of May, but in the days prior the explosive activity had increased and would increase even further in the coming days. That shift in activity marks the end of the 2nd phase and the start of the 3rd phase.