Last pair and edges of the last layer is a way to solve the last F2L pair and all edges of the last layer.
Intermediate
This is divided into two substeps:
 LPEOLL, orient all edges and pair up (any order). This is a intuitive step. Using algorithms is possible but you would need the same number as for ZBLS.
 LPEPLL, place the last pair and permute all edges. There are six cases and their mirrors.
LPELL is maybe not so useful for speedsolving, but for FMC. After this step is done you will have L4C left, 1:3 times it will be only L3C and 1:324 you will have a complete LLskip. All L4C cases are easy to solve using one or two commutators. In FMC, to save moves the commutators are preferably inserted in the skeleton if such a point is found.
Optimal algs for the second step are found lower at this page.
Advanced
A second way to solve this step is to first pair up and then do the rest in one look. There are 48 + 48 mirror cases for the second half. An advanced method that, if you include L4C places the last pair and solve all of the last layer in two looks and 'only' 180 algs. Recognition for the edges is awful if you just look at it, but is not harder than COLL or something, if you use sticker colour recognition.
The cases are not listed on the internet, some day you may find them here...
Mad
 All in one?
 Forget it! There are thousands of cases. (six times ZBLS)
LPEPLL Cases
Note that all of these algorithms are written in the Western notation, where a lowercase letter means a doublelayer turn and rotations are denoted by x, y, and z. (how to add algorithms)
Click on an algorithm (not the camera icon) to watch an animation of it.

The names for the cases are where two of the edges will go, if it is a R side case, then these are first the edge sitting in UL and then the one in UB. For the L side cases these are UR and UB. The images assumes the UF edge is solved if the pair is above the slot, if it is some diffrent edge than UF, then just AUF it to solved position for recognition. Some algs may need a leading AUF if you are in the same position as the image, these are not explicity written here (the animations shows the correct position).
The average number of moves is 6 HTM not including any leading or ending AUF. All cases are having the same probability (1:2 R or L and 1:6 within these groups). The algs here are all optimal, if there are more than one for a case, then the other(s) does some diffrent LLcorner case than the first one.
R side pair
R LB

Name: R LB
Used in: LPEPLL
Optimal moves: 9 HTM
All solved here, but just placing the pair will swap two edges, that are optimally solved by sneaking in a backside Antisune.


R RB

Name: R RB
Used in: LPEPLL
Optimal moves: 3 HTM
Z edges, just place from U2 position.


R LR

Name: R LR
Used in: LPEPLL
Optimal moves: 3 HTM
The usual R U' R' pair.


R RL

Name: R RL
Used in: LPEPLL
Optimal moves: 7 HTM
Unexpected conjugate to solve.


R BL

Name: R BL
Used in: LPEPLL
Optimal moves: 7 HTM
Sune style solution.


R BR

Name: R BR
Used in: LPEPLL
Optimal moves: 7 HTM
A 3cycle commutator.


L side pair
L RB

Name: L RB
Used in: LPEPLL
Optimal moves: 9 HTM
Mirror of R LB.


L LB

Name: L LB
Used in: LPEPLL
Optimal moves: 3 HTM
Mirror of R RB.


L RL

Name: L RL
Used in: LPEPLL
Optimal moves: 3 HTM
Mirror of R LR.


L LR

Name: L LR
Used in: LPEPLL
Optimal moves: 7 HTM
Mirror of R RL.


L BR

Name: L BR
Used in: LPEPLL
Optimal moves: 7 HTM
Mirror of R BL.


L BL

Name: L BL
Used in: LPEPLL
Optimal moves: 7 HTM
Mirror of R BR.

