Laboratory activities and constructivism are two notions that have been playing significant roles in science education. Despite common beliefs about the importance of laboratory activities, reviews reported inconsistent results about the effectiveness of laboratory activities. Since laboratory activities can be expensive and take more time, there is an effort to introduce virtual laboratory activities. This study aims at exploring the learning environment created by a virtual laboratory and a real laboratory. A quasi experimental study was conducted at two grade ten classes at a state high school in Bandung, Indonesia. Data were collected using a questionnaire called Constructivist Learning Environment Survey (CLES) before and after the laboratory activities. The results show that both types of laboratories can create constructivist learning environments. Each type of laboratory activity, however, may be stronger in improving certain aspects compared to the other. While a virtual laboratory is stronger in improving critical voice and personal relevance, real laboratory activities promote aspects of personal relevance, uncertainty and student negotiation. This study suggests that instead of setting one type of laboratory against the other, lessons and follow up studies should focus on how to combine both types of laboratories to support better learning.